Abstract:The motorized spindle is a key component of high-end computer numerical control (CNC) machining centers. However, motorized spindles possess the characteristics of compact structure, large heat generation, and complex thermal mechanisms, making it difficult to directly measure and estimate the temperature of internal regions. To address this issue, a real-time temperature estimation method is proposed based on the fusion of mechanism and data. First, by analyzing the structure and thermal properties of the motorized spindle, a thermal resistance network is constructed and expressed as a time-varying state-space expression with different types of unknown parameters. Then, the model parameters are optimized by combining the temperature field simulation data and the measured temperature from multiple operating conditions. Applications of monitoring the temperature of the undetectable region show that real-time temperature monitoring is crucial to the overheating protection of motorized spindles. Furthermore, by subdividing the grid and fuzzy interpolation calculation, a refined transient temperature field distribution of the motorized spindle is generated in real time. In the verification of the accuracy of indirectly estimated temperature distribution for the internal undetectable region, the average error of the estimated steady-state temperature for the undetectable region is 2.96 °C, and the maximum error is 8.65 °C. In addition, a thermal fault monitoring method was proposed to monitor faults of undetectable heat sources and has higher accuracy and faster response speed than the typical temperature threshold method. This approach integrates finite-element (FE) simulation technology, thermal resistance network modeling, and data-driven methods to achieve real-time monitoring of temperature fields in the, otherwise, inaccessible regions of the motorized spindle and has high industrial application potential in the fields of intelligent spindles and digital twins.

Online estimation of the heat flux during turning using long short-term memory based encoder-decoder

Machine Learning Strategy for Wall Heat Flux Prediction in Aerodynamic Heating

Deep-Learning Strategy Based on Convolutional Neural Network for Wall Heat Flux Prediction

Unsteady Heat Flux Measurement and Predictions Using Long Short-Term Memory Networks

Prediction of Thermal Deformation and Real-Time Error Compensation of a CNC Milling Machine in Cutting Processes

A Modeling Method for Thermal Error Prediction of CNC Machine Equipment Based on Sparrow Search Algorithm and Long Short-Term Memory Neural Network

Real-time tool condition monitoring method based on in situ temperature measurement and artificial neural network in turning

A comparison of ultrasonic temperature monitoring using machine learning and physics-based methods for high-cycle thermal fatigue monitoring

Research on the Construction Method of the MCNN-LSTM-SVM Fusion Model for Thermal Error of the Precision Boring Machine Feed Axis with Small Samples

Self learning-empowered thermal error control method of precision machine tools based on digital twin

An axial attention-BiLSTM-based method for predicting the migration of CNC machine tool spindle thermal error under varying working conditions

Digital Twin-Driven Thermal Error Prediction for CNC Machine Tool Spindle

Fast reconstruction of milling temperature field based on CNN-GRU machine learning models

Determination of tool tip steady-state temperature in dry turning process based on artificial neural network

Integrated thermal error modeling and compensation of machine tool feed system using subtraction-average-based optimizer-based CNN-GRU neural network

Input Variable Selection and Structure Optimization for LSTM-Based Soft Sensor With a Dual Nonnegative Garrote Approach

Machine Tool Wear Prediction Technology Based on Multi-Sensor Information Fusion

Learning to Monitor Machine Health with Convolutional Bi-Directional LSTM Networks

A novel method for tool condition monitoring based on long short-term memory and hidden Markov model hybrid framework in high-speed milling Ti-6Al-4V

In-process tool condition forecasting based on a deep learning method

Real-Time Temperature Estimation in the Undetectable Region of Motorized Spindle